CN1299150A

CN1299150A - Combined elements separating method, film mfg. method and combined elements separating device

Info

Publication number: CN1299150A
Application number: CN00135339A
Authority: CN
Inventors: 柳田一隆; 近江和明; 坂口清文
Original assignee: Canon Inc
Current assignee: Canon Inc
Priority date: 1999-12-08
Filing date: 2000-12-08
Publication date: 2001-06-13
Also published as: KR100433057B1; EP1107295A2; TW508690B; US6653205B2; US20010003668A1; KR20010062241A; EP1107295A3

Abstract

This invention relates to a composite member separating method in which a first member (1) having a separation layer (4) and a transfer layer (5) on the separation layer (4) is bonded to a second member (2) is separated at a position different from the bonding interface between the first member (1) and the second member (2), the method comprising the steps of, applying a force asymmetric with respect to the interface to the end portion of the composite member to form a crack (7A) that runs from the surface of the first member (1) to the separation layer (4) through the transfer layer (5), and then, growing the crack is grown along the separation layer (4) to completely separate the composite member.

Description

Combined element separating method, thin film manufacturing method, and combined element separating apparatus

The composite member relates to a composite member separation method, a thin film manufacturing method, and a composite member separation apparatus.

Various manufacturing methods have been proposed for forming a thin film of an insulator (dielectric material), a conductor, a semiconductor, or a magnetic material, including a method of forming a thin film by separating a composite member composed of at least two thin layers having different materials or different structures.

To facilitate understanding of these methods, the SOI layer, which is a thin film, will be described below.

Japanese Patent Laid-Open No.7-302899 and U.S. patent No.5,856,229 disclose methods of forming a combined element by bonding a first element having a porous layer formed on a single-crystal Si substrate and a single-crystal non-porous layer thereon to a second element via an insulating layer, and separating the combined element into two parts at the porous layer functioning as a separation layer, thereby converting the single-crystal non-porous layer to the second element. This technique is extremely advantageous because the film thickness uniformity of the SOI layer is good, the crystal defect density in the SOI layer can be reduced, the surface flatness of the SOI layer is good, and an SOI substrate having an SOI layer with a thickness of about several tens of nanometers to ten micrometers can be manufactured.

Further, this method is advantageous because the single crystal Si substrate and the second member can be separated without a large portion of the single crystal Si substrate being damaged, and the separated Si substrate can be reused.

In order to separate a bonded substrate stack into two substrates, for example, by pulling the two substrates apart when a force is applied in a direction perpendicular to the bonding interface, applying a shear force parallel to the bonding interface (for example, moving the two substrates in opposite directions in a plane parallel to the bonding interface or rotating the two substrates in opposite directions when a force is applied in a circumferential direction), applying a pressure in a direction perpendicular to the bonding interface, applying sonic energy such as ultrasonic waves to the separation region, inserting a peeling member (for example, a sharp blade such as a knife) from the edge of the side surface of the bonded substrate stack into the separation region parallel to the bonding interface, applying an expansion energy of a substance filling up the pores in the porous layer functioning as the separation region, and thermally oxidizing the porous layer functioning as the separation region from the side surface of the bonded substrate stack to expand into the volume of the porous layer to separate the substrates, or separating the substrate by selectively etching a porous layer functioning as a separation region from the side surface of the bonded substrate stack.

Such methods are disclosed in U.S. Patent No.5,854,123, Japanese Patent Laid-Open No.11-237884, 10-233352 or European Patent Laid-Open No. 0867917.

Japanese Patent Laid-Open No.5-211128 and U.S. Patent No.5,374,564 disclose methods of separating a bonded substrate stack into two parts by implanting, for example, hydrogen ions from the upper surface side thereof into a single crystal Si substrate to form an ion implantation layer serving as a separation layer in which the concentration of implanted ions is locally high in the substrate, bonding the resultant substrate to another substrate, and heating the bonded substrate stack.

In the above-described separation method, it is important how to stabilize the position of crack formation at the start of separation.

For example, if a crack formed on a portion other than the separation layer in the separation-bonded substrate stack grows toward the center of the substrate, the thin film as a future SOI layer may be broken toreduce the throughput of the SOI substrate.

The present invention has been made in view of the above circumstances, and a first object of the present invention is to improve the repeatability of crack generation sites in a combined member such as a bonded substrate stack and to suppress the amount of damage to a thin film on an end portion.

A second object of the present invention is to appropriately separate a combined element on a separation layer such as a porous layer or an ion-implanted layer.

According to a first aspect of the present invention, there is provided a combined element separating method of separating a combined element formed by bonding a first element having a separation layer and a conversion layer on the separation layer with a second element at a position different from a bonding interface between the first element and the second element, comprising a pre-separation step including a step of applying an asymmetric force with respect to the bonding interface to an end face portion of the combined element to form a crack in the combined element extending from a surface of the first element through the conversion layer to the separation layer, and an actual separation step including a step of growing the crack along the separation layer.

According to a second aspect of the present invention, there is provided a thin film manufacturing method including a step of bonding a first element and a second element having a separation layer and a conversion layer on the separation layer to form a combined element and a step of separating the combined element at a position different from a bonding interface between the first element and the second element, wherein the separating step includes a first step of forming a crack extending from a surface of the first element to the separation layer through the conversion layer in the combined element by applying an asymmetric force with respect to the bonding interface to an end face portion of the combined element and a second step of growing the crack along the separation layer.

According to a third aspect of the present invention, there is provided a composite element separating apparatus for separating a composite element formed by bonding a first element having a separation layer and a conversion layer on the separation layer with a second element at a position different from a bonding interface between the first element and the second element, comprising pre-separation mechanism for applying an asymmetric force with respect to the bonding interface to an end face portion of the composite element to form a crack in the composite element extending from a surface of the first element to the separation layer through the conversion layer, and actual separation mechanism for growing the crack along the separation layer.

Other features and advantages of the present invention will be apparent from the following description taken in conjunction with the accompanying drawings, in which like reference characters designate the same or similar elements throughout the figures thereof.

Brief description of the drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.

FIGS. 1A, 1B and 1C are schematic views for explaining a method of separating a combined member according to a preferred embodiment of the present invention;

fig. 2A is a schematic view for explaining a process of forming a porous layer in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2B is a schematic view for explaining a process of forming a single crystal Si layer and an insulating layer in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2C is a schematic view for explaining a bonding process in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2D is a schematic view for explaining the formation of a separation start part (pre-separation process) in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2E is a schematic view for explaining a separation process (actual separation process) in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2F is a schematic view for explaining a process of removing the porous layer on the second substrate side and the SOI substrate in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

fig. 2G is a schematic view for explaining a process of removing the porous layer on the side of the first substrate in the SOI substrate manufacturing method according to the preferred embodiment of the present invention;

FIG. 3 is a schematic view showing a state where symmetrical forces are applied to first and second substrates in a bonded substrate stack;

FIG. 4 is a view schematically illustrating defects that can be created by the separation process;

FIG. 5 is a diagram illustrating the forces acting on a bonded substrate stack in forming a separation initiation portionin accordance with a preferred embodiment of the present invention;

FIG. 6 is a view schematically showing a cross-section of a bonded substrate stack having separated starts made by the process equipment according to a preferred embodiment of the present invention;

FIG. 7 is a view schematically showing the construction of a first process apparatus adapted for carrying out a preliminary separation process;

FIG. 8 is an enlarged view showing a portion of the configuration shown in FIG. 7;

fig. 9 is a view schematically showing a state where a wedge is inserted to an end face portion of a bonded substrate stack;

FIG. 10 is a view showing a wedge suitable for use in an actual separation process;

FIG. 11 is a view schematically showing the construction of a second process apparatus suitable for carrying out an actual separation process;

FIG. 12 is an enlarged view showing components in the configuration shown in FIG. 11;

fig. 13 is a view schematically showing a state where a wedge is inserted to an end face portion of a bonded substrate stack;

FIG. 14 is a view schematically showing the construction of a third process apparatus adapted for carrying out the preliminary separation process;

FIG. 15 is an enlarged view showing parts in the configuration shown in FIG. 14;

fig. 16 is a view schematically showing a state where a wedge is inserted to an end face portion of a bonded substrate stack;

FIG. 17 is a view schematically showing the construction of a fourth process apparatus suitable for carrying out the preliminary separation process;

FIG. 18 is an enlarged view showing parts inthe configuration shown in FIG. 17;

fig. 19 is a view schematically showing a state where a wedge is inserted to an end face portion of a bonded substrate stack;

FIG. 20 is a view schematically showing the construction of a fifth process apparatus adapted for carrying out an actual separation process;

FIG. 21 is a view schematically showing the construction of a sixth process apparatus suitable for carrying out a preliminary separation process or an actual separation process;

FIG. 22 is a view schematically showing the construction of a seventh process apparatus adapted to carry out a preliminary separation process or an actual separation process;

FIG. 23 is a view schematically showing the construction of an eighth process apparatus suitable for carrying out a preliminary separation process or an actual separation process;

FIG. 24 is a schematic plan view showing an example of the first substrate after separation of the combined elements according to the embodiment of the invention; and

FIGS. 25A, 25B, 25C, 25D and 25E are schematic views for explaining a thin film manufacturing method according to a preferred embodiment of the present invention.

Detailed description of the preferred embodiments

Referring to the drawings, preferred embodiments of the present invention will be described below.

Fig. 1A to 1C are schematic views for schematically illustrating a method of separating a combined member according to a preferred embodiment of the present invention.

The first element 1 has a separation layer 4 formed therein. The surface layer (conversion layer)5 in the first element is above the separation layer 4. This surface layer is later transferred to the second component. The substrate 3 of the first element is below the separation layer 4. The composite element 6 is prepared by bonding the first element 1 with the separation layer 4 inside with the second element 2. Reference numeral 1A denotes a bonding interface.

The separation method of the present embodiment includes a pre-separation process shown in fig. 1B and an actual separation process shown in fig. 1C for separating the combined member 6 at a position different from the bonding interface 1A between the first member 1 and the second member 2.

During the pre-separation process, as shown in fig. 1B, an asymmetric force 8 is applied to the end face portion of the combination element 6 with respect to the bonding interface 1A to form a crack 7A in the combination element 6 extending from the surface of the first element 1 through the conversion layer 5 to the separation layer 4. Such an asymmetric external force 8 with respect to the bonding interface suppresses the crack 7A from extending horizontally parallel to the bonding interface toward the center of the combined element 6. In the example shown in fig. 1B, the end of the crack 7A reaches the upper surface of the substrate 3. However, the end of the crack 7A only needs to reach the upper surface or the inside of the separation layer 4.

During the actual separation process, the crack 7B grows along the separation layer 4 from the separation-start portion where the crack 7A has been formed, as shown in fig. 1C. The crack 7A has reached the separation layer 4 in the first component 1 during the pre-separation process. Then, a separating force 9 separating the first and second elements is applied so that a crack 7B is generated in the separation layer 4 or an interface between the upper and lower positions of the separation layer 4 and grows along the separation layer 4. Fig. 1C illustrates a situation in which a crack 7A grows within the separation layer 4.

Thus, a very large part of the conversion layer 5 is converted to the second element without damage. Only a small part of the end portion of the conversion layer 5 that is not converted into the second element is lost.

The conversion layer 5 used in the present invention means a thin layer which forms cracks during the pre-separation process and is converted to the side of the second element by separation during the actual separation process. As the conversion layer 5, at least one thin layer selected from thin layers consisting of an insulator, a conductor, a semiconductor, and a magnetic material is used. Examples of the semiconductor thin layer are a single crystal Si layer, a polycrystalline silicon layer, an amorphous Si layer, a Ge layer, a SiGe layer, a SiC layer, a C layer, and a compound semiconductor (e.g., GaAs, InP, or GaN) layer. Examples of the insulator are a silicon oxide layer, a silicon nitride layer, and a silicon oxynitride layer. When ion implantation (described later) is used as a method of forming the separation layer, the surface of the implanted substrate is used as a conversion layer.

The separation layer 4 used in the present invention is a thin layer constituting the exposed surface after separation according to the formation of cracks in the separation layer or at the interface above or below during the actual separation process. Among some elements of the composite element, a layer region or interface of a structure having a relatively low mechanical strength or a layer region or interface where stress is relatively concentrated can be used as a separation layer in the present invention. More specifically, a porous layer formed by anodizing or implanting ions of hydrogen, nitrogen, or a rare gas can be employed. When ion implantation is performed, stress ordefects are generated in the structure while reducing the mechanical strength. For this reason, such a structure can be used as a separation layer even when the non-porous layer is composed of micro-cavities.

Porous Si was found in 1956 by uhlri et al, who studied semiconductor electropolishing (a. uhlrill bell syst. tech. j. vol 35,333 (1956)). Porous Si can be formed by anodizing a Si substrate in an HF solution.

Unnagami et al studied the dissolution reaction of Si at the time of anodizing and reported that holes are necessary for the anodizing reaction of Si in an HF solution, and are the reactions described below (T. Unnagami, J. electrochem. Soc., Vol.127,476 (1980)).

Or

In the formula e⁺And e^-Respectively, holes and electrons, and n and λ represent the number of holes necessary to dissolve one Si atom. Accordingly, when n>2 or λ>4, porous Si is formed. The above fact suggests that P-type Si having voids is converted into porous Si while n-type is not converted. Nagano et al and Imai have reported selectivity in this transition (Nagano. Nakajima, Anno, Onaka and Kajiwara, IEICE Technical Report, Vol79, SSD79-9549(1979)), (K.Imai, Solid-State Electronics, Vol.24,159 (1981)).

However, n-type Si has also been reported to be converted to porous Si at high concentrations (r.p. holmstrom and j.y.chi, appl.phy.lett., Vol 42,386 (1983)). Therefore, it is important to select a substrate that can be converted into a porous Si substrate without depending on p-type or n-type.

It is also preferable to prepare the composite member by bonding the first and second members, and annealing the composite member in an oxidizing atmosphere to increase the bonding strength to form an oxide film on the exposed surface of the composite member. In such a case, since the end face portion of the composite member serving as the separation start portion is covered with the oxide film, satisfactory separation can be performed by the method of the present invention.

Referring to the drawings, how to apply an asymmetric external force used during the pre-separation process and how to apply an external force used during the actual separation process will be described below.

Fig. 2A to 2G are schematic diagrams for schematically illustrating a method of manufacturing an SOI substrate according to an embodiment of the present invention.

In the step shown in fig. 2A, a single crystal Si substrate 11 is prepared, and a porous Si layer (separation layer) 12 is formed on the surface of the single crystal Si substrate 11 by, for example, anodizing treatment.

In the step shown in FIG. 2B, the porous layer is grown by epitaxyA non-porous single crystal Si layer 13 is formed on the Si layer. The surface of the non-porous single crystal Si layer 13 is oxidized to form an insulating layer (SiO)₂Layer) 14. With such a process, a first substrate (first element) 10 is formed. The single crystal Si layer 13 and the insulating layer 14 constitute a conversion layer.

The porous Si layer 12 can be formed by, for example, a method of implanting hydrogen ions or ions of an inert gas such as helium into the single crystal Si substrate 11 (ion implantation method). The porous Si layer formed by this method has a large number of microporosities and is therefore referred to as a microporosity layer.

In the step shown in fig. 2C, a second substrate (second element) 20 composed of single crystal Si is prepared and brought into close contact with the first substrate 10 at room temperature while the insulating layer 14 is made to oppose the second substrate 20, thereby forming a bonded substrate stack (combined element).

In view of the state shown in fig. 2C obtained when the first and second substrates are brought into close contact with each other, as described above, the insulating layer 14 may be formed on the side of the single crystal Si layer 13 of the first substrate, on the second substrate, or on both the single crystal Si layer 13 and the second substrate 20. However, when the insulating region 14 is formed on the side of the single crystal Si layer 13 serving as the active layer, the bonding interface (bonding interface) between the first and

second substrates

10 and 20 is independent of the active layer. For this reason, an SOI substrate having higher quality can be obtained.

In the step (pre-separation process) shown in fig. 2D, a separation start portion 60 is formed as a portion where separation is started. The separation of the bonded substrate stack 30 is started in the next actual separation process (fig. 1E) from the separation start portion 60.

In the preferred embodiment of the present invention, an asymmetric force is applied to the end face portion of the bonded substrate stack 30 relative to the bonding interface between the first and

second substrates

10 and 20. In particular, stripping elements with an asymmetric structure (e.g., solid wedges or wedges) or substrate clamping mechanisms with an asymmetric structure are employed.

In order to apply "asymmetric force" to the first substrate 10 and the second substrate 20, for example, the following method is preferably employed.

A solid wedge having a structure and/or function that applies an asymmetric force to the bonding interface between the first and

second substrates

10 and 20 is preferably inserted into the end face portion of the bonded substrate stack 30.

More specifically, for example, a wedge having an asymmetrical shape with respect to the bonding interface between the first and

second substrates

10 and 20 is inserted to the end face portion of the bonded substrate stack 30.

Another method is to insert a wedge of varying hardness between the portion against the first substrate 10 and the portion against the second substrate 20 to the end face portion of the bonded substrate stack 30.

In addition, a symmetric or asymmetric wedge is inserted into the bonding interface between the first and

second substrates

10 and 20 and the wedge is vibrated in a direction not perpendicular to the bonding interface (in the direction of asymmetry).

As still another method of applying an "asymmetric force" to the first substrate 10 and the second substrate 20, for example, a substrate holding portion for holding the

substrates

10 and 20 is formed in an asymmetric structure to change a degree of warpage at an end portion between the substrates. When the end portion of the first substrate 10 is warped by receiving a bending moment larger than that of the end portion of the second substrate, a desired crack can be generated.

The relation between the first force acting on the first substrate 10 and the second force acting on the second substrate 10 is determined such that the appropriate separation initiation 60 is formed under the condition that the first force is made different from the second force.

The proper separation initiation means a separation initiation constituting a structure for separating the bonded substrate stack 30 into two substrates by almost exclusively rupturing the porous layer in the subsequent actual separation process (fig. 2E).

More specifically, the appropriate separation start portion has a structure that partially exposes the porous layer to the atmosphere outside the bonded substrate stack 30, and can be formed by generating a crack 7A that extends from the surface of the first substrate 10, i.e., the surface of the insulating layer 14 in this case, to the porous layer via the insulating layer 14 and the single crystal Si layer 13.

Although fig. 2D shows the separation starting portion 60 formed after the removal of the portion on the end (outside) of the crack 7A or simultaneously with the formation of the crack 7A, it is not always necessary to remove the portion on the end (outside) of the crack 7A as shown in fig. 1B.

As described above, in forming the separation start portion 60, the porous layer 12 can be selectively broken during the subsequent actual separation process to separate the bonded substrate stack 30 into two substrates, and thus defects can be effectively prevented from being generated during the separation process.

Next, in the step shown in fig. 2E (actual separation process), the bonded substrate stack 30 having the separation start portion 60 is completely separated into two substrates at the porous layer 12 portion. The bonded substrate stack 30 is separated by growing a crack horizontally along the porous layer 12 starting from the separation start portion 60. For the separation, for example, the following method is preferably employed.

(1) Separation processes using liquids

A jet of fluid(e.g., a liquid such as water or a gas such as air or nitrogen) is sprayed onto bonded substrate stack 30 so that porous layer 12 near the beginning portion 60 is broken, whereupon the fluid is sprayed onto the surface to gradually grow a separation region (separation zone), i.e., a horizontal crack. When the bonded substrate stack 30 having the separation start portion 60 is rotated while jetting the fluid toward the end surface portion of the bonded substrate stack 30, the crack growth is simultaneously started when the fluid rushes toward the separation start portion 60. Thus, at the start of the actual separation process, there is no need to position the fluid relative to the separation initiation site 60.

Thereafter, as the separation progresses, the remaining porous layer 12 is entirely broken, thus completely separating the bonded substrate stack 30. At this time, it is preferable that the separation is performed while changing the position of ejecting the fluid to the bonded substrate stack 30 by rotating the bonded substrate stack 30 in its plane.

When a fluid is used, a method using a so-called static pressure fluid in which a sealed space around a separation start point is formed with a sealing member such as an O-ring, the space is filled with the fluid, and the fluid is pressurized to grow a crack and to inject the fluid to the crack may be employed instead of the above-described method using a jet of fluid.

(2) Separation method using solid wedges

The bonded substrate stack 30 is completely separated by gradually inserting a wedge (e.g., a thin wedge composed of resin) into the separation start portion 60 in the bonded substrate stack 30 to separate the two substrates from each other.

When the separation-starting portion 60 is formed by the wedge, two processes, i.e., the process of forming the separation-starting portion 60 (fig.2D) and the actual separation process (fig. 2E), can be performed in succession using a single apparatus.

(3) Separation method by peeling

One surface of the bonded substrate stack 30 is fixed, and the other surface is pulled in the axial direction of the bonded substrate stack 30 with a flexible tape or the like, thereby completely separating the bonded substrate stack 30 at the porous layer 12 portion. At the start of separation, a pulling force is applied so that the separation force is concentrated to the separation initiation portion 60.

(4) Separation method using shear stress

One surface of the bonded substrate stack 30 is fixed, and a force is applied to the other surface to move it in the planar direction of the bonded substrate stack 30, whereby the shear stress completely separates the bonded substrate stack 30 at the porous layer 12 portion. For such methods, it is preferred that the separation initiation portion 60 be relatively large compared to other methods so that separation begins from the separation initiation portion 60.

As described above, when the separation start portion 60 is formed and then the separation operation (actual separation operation) is started from the separation start portion, the bonded substrate stack 30 can be separated almost only at the porous layer portion. Therefore, the interface between the single crystal Si layer 13, the insulating layer 14, the second substrate 20, the single crystal Si substrate 11, and these thin layers or substrates can be prevented from being broken and serious defects can be prevented from being generated.

Compared with the preferred embodiment of the present invention, the separation condition of the bonded substrate stack 30 when the separation starting portion is formed by applying the symmetrical force to the first substrate 10 and the second substrate 20 and when the bonded substrate stack 30 is separated without forming the separation starting portion by applying the symmetrical force to the first substrate 10 and the second substrate 20 will be explained.

Fig. 3 is a schematic view showing a state where symmetrical forces are applied to the first substrate 10 and the second substrate 20 in the bonded substrate stack.

As shown in fig. 3, when

forces

902A and 902B symmetrical with respect to the bonding interface between the first and

second substrates

10 and 20 are applied to the first substrate 10 and the second substrate 20, respectively,

tensile forces

903A and 903B symmetrical with respect to the bonding interface act on an action point 901 at the outermost portion of the bonding interface. Thus, the direction of the separating operation is almost a planar direction (horizontal direction in fig. 3) as indicated by an arrow 904 in the bonded substrate stack 30.

Therefore, as indicated by an arrow 905 shown in fig. 4, the single-crystal Si layer 13 and the insulating layer (SiO) that are fragile next to the porous layer 12 only along the structure are separated₂Layer) 14 does not reach porous layer 12 but only by a suitable distance L. In this case, the single crystal Si layer 13 and the insulating layer (SiO) are not formed along the silicon layer₂Layer) 14 to the side of the second substrate 20, a portion corresponding to the length L and a portion of the outer surface of the portion are left out to cause defects.

However, since the single crystal Si layer 13 and the insulating layer (SiO)₂Layers) 14 have a ratio of interface to interfaceThe porous layer 12 has a high structural strength, so that horizontal cracks generated in the interface rarely reach the center of the substrate. Instead, the cracks reach porous layer 12through the middle of single crystal Si layer 13, and thus are mostly separated on porous layer 12. If distance L is long, the distance between the point at which the crack reaches porous layer 12 and the outer edge portion of the substrate may exceed 3 mm.

In contrast, according to the preferred embodiment of the present invention, as shown in FIG. 5, forces 906A and 906B asymmetrical with respect to the bonding interface between the first and

second substrates

10 and 20 are applied to the portions where the separation initiation portions should be formed, thereby applying asymmetrical forces 907A and 907B to the bonding interface between the first and

second substrates

10 and 20.

Thus, as indicated by arrow 908, the crack can easily reach porous layer 12.

Fig. 6 is a view showing a cross-section of a combined member after a preliminary separation process according to an embodiment of the present invention. Fig. 6 shows a case in which a crack reaches porous layer 12 from action point 901 on the first element surface across insulating layer 14 and single crystal Si layer 13. The portion 61 on the end of the crack (on the outside) may be removed before the actual separation process. According to the embodiment of the present invention, the distance (L in fig. 4) by which the crack extends along the interface between the single crystal Si layer 13 and the insulating layer 14 can be made zero or very short. More specifically, the distance between the outer edge portion of the substrate and the point at which the crack reaches the porous layer can be 3 mm or less, and more preferably 2 mm or less.

During the separation process shown in fig. 2E (actual separation process), the first substrate 10' after separation has a porous layer 12b on the single crystal Si substrate 11. On the other hand, the second substrate 20' after the separation has a multilayer structure of the porous Si layer 12 c/the single crystal Si layer 13 b/the insulating layer 14b, and the single crystal Si substrate 20.

As for the above-described processes, the single crystal Si layer 13 and the insulating layer 14 on the porous layer 12 in the first substrate can be transferred to the second substrate. Porous layer 12 is an example of a separation layer, and single-crystal Si layer 13 and insulating layer 14 are an example of a conversion layer (surface layer) that converts from a first substrate to a second substrate.

In the step shown in fig. 2F, porous layer 12c on the surface of second substrate 20' after separation is selectively removed because of necessity. Thus, as shown in fig. 2F, an SOI substrate 50 having a multilayer structure of the single crystal Si layer 13 b/the insulating layer 14 b/the single crystal Si substrate 20, i.e., an SOI layer (thin film) 13b, is obtained.

In the step shown in fig. 2G, the porous layer 12b on the single crystal Si substrate 11 in the first substrate 10' after separation is selectively removed by etching or the like, as necessary. The single crystal Si substrate 11 thus obtained can be reused as a substrate for forming the first substrate 10 or as the second substrate 20.

As described above, according to the preferred embodiment of the present invention, the separation-starting portion having the crack reaching the separation layer is formed by applying the force asymmetric with respect to the bonding interface between the first and second substrates to the end face portion of the bonded substrate stack, and the separation is completed from the separation-starting portion. Since the porous layer can be selectively broken, any serious defect can be prevented.

Next, a process apparatus suitable for carrying out a process (pre-separation process) for forming a separation start part in the present invention will be described.

[ first Process Equipment]

FIG. 7 is a view schematically showing the construction of a first process apparatus adapted to carry out a preliminary separation process. The process apparatus 200 shown in fig. 7 has a base table 201 having a clamping portion 203 for fixing the bonded substrate stack 30, an elastic body 202 for pressing the bonded substrate stack 30 against the clamping portion 203, a wedge 210, a driving shaft 211 having a rack for reciprocally moving the wedge 210, a guide member 212 for guiding the driving shaft 211, and a motor (driving member) 213 having a pinion for applying a driving force to the driving shaft 211 and reciprocally moving the driving shaft 211.

To complete the pre-separation process, motor 213 is rotated in a positive direction so that wedges 210 are inserted only a predetermined amount into the end face portion of bonded substrate stack 30. Conversely, to withdraw wedge 210, motor 213 is rotated in the opposite direction.

Fig. 8 is an enlarged view showing parts in the configuration shown in fig. 7. The wedge 210 having an asymmetric structure has two contact surfaces having different inclination angles with respect to the bonding interface (the inclination angle θ 1 on the side abutting the first substrate 10 is larger than the inclination angle θ 2 on the side abutting the second substrate 20) as a structure for applying asymmetric force with respect to the bonding interface between the first and

second substrates

10 and 20 to the first and

second substrates

10 and 20, respectively. Such an asymmetric structure can generate a crack from the exposed portion of the first substrate 10 toward the inside of the porous layer, and form an appropriate separation start portion.

Fig. 9 is a view schematically showing a state where thewedge 210 is inserted into the end face portion in the bonded substrate stack 30. Since the wedge 210 has an asymmetric structure, the first substrate 10 and the second substrate 20 are subjected to asymmetric force with respect to the bonding interface. Fig. 9 macroscopically illustrates cracks on the separation initiation. In fact, as mentioned above, cracks include cracks that reach the separation layer through the conversion layer and horizontal cracks that occur within the separation layer or in interfaces on the upper or lower side of the separation layer.

In the actual separation process, as shown in fig. 10, it is preferable to use a solid wedge 210' having a symmetrical structure. In such a case, the process equipment shown in fig. 7 is used for the pre-separation process, and the process equipment shown in fig. 10 (process equipment having a wedge 210' instead of the wedge 210 in the process equipment shown in fig. 7) is used for the actual separation process.

[ second Process Equipment]

FIG. 11 is a view schematically showing the construction of a second process apparatus adapted to carry out a preliminary separation process. The process apparatus 300 shown in fig. 11 has a base table 201 having a clamping portion 203 for fixing the bonded substrate stack 30, an elastic body 202 for pressing the bonded substrate stack 30 against the clamping portion 203, a wedge 220, a driving shaft 211 having a rack for reciprocally moving the wedge 220, a guide member 212 for guiding the driving shaft 211, and a motor 213 having a pinion for applying a driving force to the driving shaft 211 and reciprocally moving the driving shaft 211.

To complete the pre-separation process, motor 213 is rotated in a positive direction so that wedges 220 are inserted only a predetermined amount into the end face portion of bonded substrate stack 30. Conversely, to withdraw wedge 220, motor 213is rotated in the opposite direction.

Fig. 12 is an enlarged view showing parts in the configuration shown in fig. 11. The wedges 220 having an asymmetric structure have two contact surfaces against the first and

second substrates

10 and 20, respectively, as a structure for applying an asymmetric force to the first and

second substrates

10 and 20 with respect to the bonding interface between the first and

second substrates

10 and 20. The second contact surface against the second substrate 20 is set back a length D from the first contact surface against the first substrate. Such an asymmetric structure can generate a crack from the exposed portion of the first substrate 10 toward the inside of the porous layer to form an appropriate separation start portion on the end face portion of the bonded substrate stack 30.

Fig. 13 is a view schematically showing a state where the wedge 220 is inserted into the end face portion in the bonded substrate stack 30. Since the wedges 220 have an asymmetric structure, the first substrate 10 and the second substrate 20 are subjected to asymmetric forces with respect to the bonding interface. Fig. 13 macroscopically illustrates cracks at the beginning of separation. In practice, cracks include cracks that reach the separation layer through the conversion layer and horizontal cracks that occur within the separation layer or in the interface of the upper or lower side of the separation layer.

In the actual separation process, as shown in fig. 10, it is preferable to use a solid wedge 210' having a symmetrical structure. In such a case, the process equipment shown in fig. 11 is used for the pre-separation process, and the process equipment shown in fig. 10 (process equipment having a wedge 210' instead of the wedge 220 in the process equipment shown in fig. 11) is used for the actual separation process.

[ third Process Equipment]

FIG. 14 is a view schematically showing the construction of a third process apparatus adapted to carry out a preliminary separation process. The process apparatus 400 shown in fig. 14 has a base table 201 having a clamping portion 203 for fixing the bonded substrate stack 30, an elastic body 202 for pressing the bonded substrate stack 30 against the clamping portion 203, a wedge 230, a driving shaft 211 having a rack for reciprocally moving the wedge 230, a guide member 212 for guiding the driving shaft 211, and a motor 213 having a pinion for applying a driving force to the driving shaft 211 and reciprocally moving the driving shaft 211.

To complete the pre-separation process, motor 213 is rotated in a positive direction so that wedges 230 are inserted only a predetermined amount into the end face portion of bonded substrate stack 30. Conversely, to withdraw wedge 230, motor 213 is rotated in the opposite direction.

Fig. 15 is an enlarged view showing parts in the configuration shown in fig. 14. The wedges 230 having an asymmetric structure have a structure for applying asymmetric force to the first substrate 10 and the second substrate 20 with respect to the bonding interface between the first and

second substrates

10 and 20. More precisely, the wedge 230 has a first contact part 230a abutting against the first substrate 10 and a second contact part 230b abutting against the second substrate 20. The first contact member 230a has a hardness higher than that of the second contact member 230 b. Generally, for example, the first contact member 230a is composed of a rigid body, and the second contact member 230b is composed of an elastic body (e.g., rubber). Thus, an appropriate separation start portion can be formed on the end face portion in the bonded substrate stack 30.

Fig. 16 is a view schematically showing a state where the wedge 230 is inserted into the end face portion in the bonded substrate stack 30. Since the wedges 230 have an asymmetric structure, the first substrate 10 and the second substrate 20 are subjected to asymmetric forces with respect to the bonding interface. Fig. 16 macroscopically illustrates cracks at the beginning of separation. In fact, as mentioned above, cracks include cracks that reach the separation layer through the conversion layer and horizontal cracks that occur within the separation layer or in interfaces on the upper or lower side of the separation layer.

In the actual separation process, as shown in fig. 10, it is preferable to use a solid wedge 210' having a symmetrical structure. In such a case, the process equipment shown in fig. 14 is used for the pre-separation process, and the process equipment shown in fig. 10 (process equipment having a wedge 210' instead of the wedge 230 in the process equipment shown in fig. 14) is used for the actual separation process.

[ fourth Process Equipment]

FIG. 17 is a view schematically showing the construction of a fourth process apparatus adapted to carry out the preliminary separation process. The process apparatus 500 shown in fig. 17 has a base table 201 having a clamping portion 203 for fixing the bonded substrate stack 30, an elastic body 202 for pressing the bonded substrate stack 30 against the clamping portion 203, a wedge 240, a driving shaft 211 'having a rack for reciprocally moving the wedge 240, a guide member 212 for guiding the driving shaft 211', a motor 213 having a pinion for applying a driving force to the driving shaft 211 'and reciprocally moving the driving shaft 211', a vibration member 250 for vibrating the wedge 240 in a direction asymmetrical with respect to the bonding interface between the first and

second substrates

10 and 20, and a coupling member 241 for coupling one end of the vibration member 250 with the wedge 240.

The vibration element 250 generates vibration in a direction asymmetrical with respect to the bonding interface between the first and

second substrates

10 and 20. As the vibration element 250, for example, an element (for example, a piezoelectric element) that converts an electric signal into mechanical vibration energy is preferably used.

To complete the pre-separation process, motor 213 is rotated in a positive direction so that wedges 240 are inserted only a predetermined amount into the end face portion of bonded substrate stack 30. Conversely, to withdraw wedge 240, motor 213 is rotated in the opposite direction.

Fig. 18 is an enlarged view showing parts in the configuration shown in fig. 17. The wedge 240 having the asymmetric structure receives vibration in a direction asymmetric with respect to the bonding interface between the first substrate 10 and the second substrate 20 from the vibration element 250 shown in fig. 17. Thus, a force asymmetrical with respect to the bonding interface is applied to the first and

second substrates

10 and 20. More specifically, the first substrate 10 comes into contact with the wedges 240 due to the above-described vibration on the portion outside the contact portion of the second substrate 20. For this reason, the moment to which the substrate 10 is subjected is larger than the moment to which the substrate 20 is subjected, so that the amount of warp on the end portion of the substrate 10 is larger than the amount of warp on the end portion of the substrate 20. This enables generation of a crack which is formed from the end portion in the exposed portion of the first substrate 10 toward the inside of the porous layer to form an appropriate separation start portion on the end face portion of the bonded substrate stack 30.

Fig. 19 is a view schematically showing a state where a wedge 240 is inserted to the end face portion of the bonded substrate stack 30. Since the wedges 240 vibrate in the direction of asymmetry, the first and

second substrates

10 and 20 are subjected to forces that are asymmetric with respect to the bonding interface. FIG. 19 macroscopically illustrates cracks at the start of separation. In fact, as mentioned above, cracks include cracks that reach the separation layer through the conversion layer and horizontal cracks that occur within the separation layer or in interfaces on the upper or lower side of the separation layer.

Such an apparatus can be suitably used even during an actual separation process when the vibration element 250 is stopped.

[ fifth Process Equipment]

Fig. 20 is a view schematically showing the construction of fifth process equipment suitable for carrying out the actual separation process of the present invention. The process apparatus 600 has a pair of

substrate holding portions

103 and 104 for fixing the bonded substrate stack 30, a rotation shaft 101 coupled to one substrate holding portion 103 and rotatably axially supporting the substrate holding portion 103, a regulator (e.g., a cylinder) 108 coupled to the rotation shaft 101 for regulating a distance between the substrate holding portion 103 and the substrate holding portion 104 to press the substrate stack 30, a rotation shaft 102 coupled to the other substrate holding portion 104 and rotatably axially supporting the substrate holding portion 104, a rotation source (motor) 105 for rotating the rotation shaft 102, a nozzle 106 for injecting a fluid into the bonded substrate stack 30, and an automatic driving device 107 for adjusting a relative positional relationship between the bonded substrate stack 30 and the nozzle 106.

As fluid, for example, a liquid like water or a gas like air or nitrogen can be used. Devices using water as a fluid are generally referred to as water jet devices. The nozzle 106 preferablyhas a diameter of about 0.1 mm, for example.

The nozzle 106 is opposed to the bonded substrate stack, and the automatic driving device 107 is driven to place the nozzle 106 at a position almost directly above the bonding interface between the first and

second substrates

10 and 20 or at a position biased to the side of the first substrate 10. In such a state, a fluid is ejected from the nozzle 106 so as to apply an external force, which is symmetrical or asymmetrical with respect to the bonding interface between the first and

second substrates

10 and 20, to the end face portion of the bonded substrate stack 30. Thus, cracks can be formed and grown in the porous layer.

Another method of applying an asymmetric force to the end face portion of the bonded substrate stack 30 relative to the bonding interface may be employed. For example, in the example shown in FIG. 20, fluid is ejected parallel to the bonding interface. However, even when the jet is ejected toward the first substrate with the jet being inclined with respect to the bonding interface, cracks can be formed and grown in the porous layer.

[ sixth Process Equipment]

Fig. 21 is a view schematically showing the configuration of a sixth process device adapted to carry out a preliminary separation process or an actual separation process.

The process apparatus 700 shown in fig. 21 has a pair of substrate holding portions 113 and 114 for rotatably holding a bonded substrate stack and a substrate holding mechanism having an asymmetric structure with a warp-restraining element 115 coupled to one substrate holding portion 114. Such a substrate clamping mechanism with an asymmetric structure applies an asymmetric force to the end face portion of the bonded substrate stack relative to the bonding interface.

The substrate 20 is sucked bythe vacuum chuck through the substrate holding portion 114 via the warp restraining member 115. The substrate 10 is sucked by a vacuum chuck through the substrate holding portion 113. When the substrate holding portion 114 (or the substrate holding portion 113) is rotated, the bonded substrate stack and the substrate holding portion 113 (or the substrate holding portion 114) are rotated together.

The process tool 700 is used in the following manner. First, the bonded substrate stack is inserted and held between the substrate holding portions 113 and 114. The bonded substrate stack is not rotated but is in a stationary state, and a fluid 117 such as pure water, air, or nitrogen is sprayed to a portion at or near the bonding interface of the bonded substrate stack. At this time, the second substrate 20 is prevented from warping upward in fig. 21 by the flat plate-like warp restraining member 115. On the other hand, on the side of the first substrate 10, since the warp members are not restrained, the outermost edges of the members holding the substrates are disposed inside the outermost edges of the second substrate 20. Therefore, the end portion of the first substrate 10 is easily warped as compared with the second substrate 20.

In such a situation, when the fluid 117 is ejected to the concave portion formed by the chamfered portions of the first and second substrates, the end face portions of the two substrates receive a separating force from the fluid. During the application of the force as shown in fig. 5, a crack forms from the first substrate surface to the separation layer as shown in fig. 6 or the like.

Next, the fluid 117 is continuously ejected while rotating the substrate holding portion 114. Thus, the rotating bonded substrate stack is gradually separated by the fluid 117. The pressure in the compressor for delivering fluid to the nozzle 106 is adjusted to completely separate the bonded substrate stack at a time scheduled for the bonded substrate stack to complete one or more revolutions. During such a separation process, cracks are spirally formed from the outer edge to the center of the substrate.

When the two substrates are completely separated, the ejection of the fluid is stopped.

[ seventh embodiment]

Fig. 22 is a view schematically showing the construction of a seventh process apparatus adapted to carry out a preliminary separation process or an actual separation process.

The process apparatus 800 shown in fig. 22 has a pair of substrate holding portions 113 and 114 for rotatably holding the bonded substrate stack and an asymmetric substrate holding mechanism having a dished warp-inhibiting element 116 for holding the substrate 20 only at the end portions. Reference numeral 118 denotes a concave portion (gap) 118. A substrate clamping mechanism having an asymmetric structure applies a force to the end face portion of the bonded substrate stack that is asymmetric with respect to the bonding interface.

The substrate 20 is held on the edge portion of the holding member by a vacuum chuck. The substrate 10 is held by the substrate holding portion 113 with a vacuum chuck. When the substrate holding portion 114 (or the substrate holding portion 113) is rotated, the bonded substrate stack and the substrate holding portion 113 (or the substrate holding portion 114) are rotated together.

The operation of this apparatus is the same as that of the apparatus shown in fig. 21.

[ eighth Process Equipment]

Fig. 23 is a view schematically showing the construction of an eighth process apparatus adapted to carry out a preliminary separation process or an actual process. The processing tool 900 has an asymmetric substrateclamping mechanism. A substrate clamping mechanism having an asymmetric configuration applies an asymmetric force to the end face portion of the bonded substrate stack relative to the bonding interface.

The process apparatus 900 shown in fig. 23 has a lower lid 121 and a lower lid 122 as well as a gap 123 and a further gap 124 for accommodating a bonded substrate stack therein.

Fluid delivery conduit 125 communicates with one gap 123 and fluid delivery conduit 126 communicates with the other gap 124 for independent delivery at a desired pressure. A sheet-like diaphragm seal member 127 composed of an elastic body such as rubber is installed inside the process equipment 900, thereby separating the gap 123 from the gap 124. A hollow O-ring 128 is disposed in the gap 124 on the side of the upper cover 122. The sealing member 127 also functions as a holding portion of the substrate 10.

The lower cover 121 has a substrate holding portion 129 for holding the substrate 20. The substrate holding portion 129 has a number of suction holes for vacuum chucks. An O-ring 131 for vacuum sealing is mounted on the peripheral portion of the substrate holding portion 129.

The process equipment 900 is used as follows. First, the upper and

lower covers

122 and 121 are opened, and the bonded substrate stack is then placed on the substrate holding portion 129 and held by suction through the suction holes 130. Next, the upper and

lower caps

122 and 121 are capped so as to hermetically seal the gap 123, and a pressurized fluid is supplied from the fluid supply pipe 125 into the gap 123. At the same time, pressurized fluid is delivered into gap 124 from fluid delivery conduit 126.

The pressure in gap 123 is made higher than the pressure in gap 124. The upper substrate 10 in the bonded substrate stack is vertically movably brought into close contact with the sealing member 127. For this reason, when the end face portion of the substrate 10 is subjected to static pressure from the fluid, the end portion of the substrate 10 is warped upward by the force action as shown in fig. 5. Thus, as shown in fig. 6 or the like, cracks are generated from the surface of the first substrate 10 to the separation layer.

Thereafter, the bonded substrate stack is completely separated while continuously applying a static pressure with the fluid.

Fig. 24 is a schematic plan view showing an example of the first substrate after separation of the combined elements according to the embodiment of the present invention.

The porous layer 126 having a uniform thickness is spread over a wide area on the separation surface side of the first substrate 11 except for the separation starting portion 60. When the distance between the position of the hemispherical separation from the initial portion and the outer edge portion of the substrate is 3 mm or less, and preferably 2 mm or less, a satisfactory SOI substrate can be obtained.

According to the embodiment of the present invention, cracks reaching the separation layer can be stably generated at the separated position of 3 mm or less, and preferably 2 mm or less from the outer edge portion of the substrate.

In the present invention, it is most preferable to perform the preliminary separation process without rotating the bonded substrate stack to prevent an undesirable increase in the area of the separation start portion 60.

After the pre-separation process, the substrate is rotated while fluid is ejected toward the end face portion of the bonded substrate stack. The crack does not grow until the fluid impacts the part near the separation start 60. When the fluid impinges on the part near the separation initiation 60, the process is switched to the actual separation process. The actual separation process using the fluid is preferably accomplished by spirally propagating a crack from the outer edge side to the center during one or more turns, preferably two or more turns, of the bonded substrate stack 30 when the rotary source 105 is driven. This suppresses warpage of the substrate during the actual separation process to thoroughly prevent the substrate itself from being broken.

Referring to fig. 25A to 25E, an embodiment of a thin film manufacturing method using ion implantation will be described.

First, a substrate like a single crystal Si wafer is prepared. It is most preferable to use a substrate having a thin layer 13 of single crystal semiconductor epitaxially grown on a mirror-surface wafer.

Next, heat treatment is performed under an oxidizing atmosphere so that an insulating layer 14 such as a silicon oxide film is formed on the surface of the substrate 11. Ions are implanted into the substrate 11 through the surface of the substrate 11 having the insulating layer 14 by a linear ion implantation method or an ion immersion ion implantation method. At this time, the energy of ion implantation is controlled according to the thickness of the conversion layer so that the implanted particles have a peak concentration at a depth corresponding to the thickness of the conversion layer.

A thin layer portion provided with a high concentration of implanted particles serves as the separation layer 4 because the implanted ions generate stress or defects.

Thus, a first element including the insulating layer 14 and the single crystal semiconductor layer 13 as an conversion layer and the separation layer 4 formed by an ion implantation method below the conversion layer was manufactured (fig. 25A).

Next, a second member such as a silicate glass or a single crystal silicon thin plate is prepared, and an insulating layer such as a silicon oxide film is formed on the surface of the second member as needed.

The first and second members 20 are brought into intimate contact and bonded at room temperature. As needed, heat treatment is preferred to increase the bond strength. Thus, a combined element shown in fig. 25B is obtained.

The external force of asymmetry described above is applied to the combined element to form a crack extending from the first element surface to the separation layer 4 through the insulating layer 14 and the single crystal semiconductor layer 13 as the conversion layer (fig. 25C).

A separation force is applied in order to grow a crack 7B in the horizontal direction in fig. 25D along the separation layer 4 (fig. 25D).

Thus, as shown in fig. 25E, the combined element is completely separated, and thus the insulating layer 14 and the single crystal semiconductor layer 13 are transferred onto the second element.

In the above process, a thin film (the single crystal semiconductor layer 13) serving as an SOI layer can be formed.

The following describes some preferred examples of the present invention.

[ example 1]

In example 1, the first process equipment 200 and the fifth process equipment 600 are applied to a process of fabricating an SOI substrate.

First, in order to form the first substrate 10, a P-type (or n-type) single crystal Si substrate 11 having a resistivity of 0.01 ohm-cm was prepared and subjected to anodizing treatment in an HF solution for two stages, thereby forming a porous Si layer 12 composed of two porous layers having different properties on the substrate surface (fig. 2A).

Current density: 7 (milliamp square centimeter)

Anodizing solution: h of HF₂O∶C₂H₅OH=1∶1∶1

Treatment time: 5 (minutes)

Porous Si thickness: 4.5 (micron)

Current density: 30 (milliamp square centimeter)

Anodizing solution: h of HF₂O∶C₂H₅OH=1∶1∶1

The process time is as follows: 10 (second)

Porous Si thickness: 0.2 (micron)

Thus, a first porous layer of low porosity is formed on the surface side of the first substrate with low current in the first period, and a second porous layer having a thickness and porosity larger than those of the first porous layer is formed under the first porous layer with high current in the second period.

The thickness of the first porous layer is not limited to the above-described examples but is preferably, for example, several hundred to 0.1 μm. The thickness of the second porous layer is not limited to the above-described example, and is, the thickness of the second porous layer can be appropriately changed according to the condition of the subsequent separation process.

The above conditions are determined so that the first porous layer has a porosity suitable for forming the high-quality Si epitaxial layer 13, and the second porous layer has a porosity that causes cracks in a portion of the second porous layer that is very close to the interface with the first porous layer.

Porous layer 12 may be comprised of one thin layer or three or more thin layers.

Next, the above-described substrate was subjected to oxidation treatment under an oxygen atmosphere for 1 hour. In the case of such oxidation, the single crystal Si structure is maintained within the porous layer while the inner walls of the pores in the porous layer are covered with a thermal oxide thin film.

A 0.3 micron thick layer of single crystal Si was epitaxially grown on the porous Si layer using CDV (chemical vapor deposition) method (fig. 2B). The growth conditions were as follows. At the beginning of the epitaxial growth, the surface of the porous Si layer 12 is placed under hydrogen gas at a high temperature. Therefore, the cavities on the surface are filled with the migrated Si atoms to form a flat surface.

Source gas: SiH₂Cl₂/H₂

Gas flow rate: 0.5/180 (liter/minute)

Gas pressure: 80 (torr)

Temperature: 950 ℃ C

Growth rate: 0.30 (micron/minute)

SiO is formed by thermal oxidation to a thickness of 200 nm on the surface of the single crystal Si layer 13 grown epitaxially₂Layer 14 (fig. 2B). With such a process, the first substrate 10 is obtained.

Separately preparing an Si substrate (second substrate) 20 having the same size as that of the first substrate 10 and making the Si substrate 20 coincide with the SiO of the first substrate 10 at the time of coincidence of the center positions of the two substrates₂The surfaces of layer 14 are in intimate contact. The composite structure was heat treated at 1180 ℃ for 5 minutes under an oxidizing atmosphere. With such a process, a bonded substrate stack 30 is formed as shown in fig. 2C. In fact, by heat treatment in an oxidizing atmosphere, the surface of the combined element is covered with an oxide film.

As shown in fig. 7, the outer peripheral portion of the bonded substrate stack 30 is held by a holding portion 203 in the process apparatus 200. At this time, the separation initiation portion formed in the bonded substrate stack 30 is made to oppose the solid wedge 210.

The motor 213 is driven to move the wedge 210 parallel to the bonding interface of the bonded stack 30 and to insert the wedge 210 into the bonded stack 30 only about 1.5 mm apart. Thus, cracks are formed in the bonded substrate stack 30 that reach the porous layer 12 as shown in fig. 6. The portion outside the crack is deleted to form a separation starting portion having a nearly semicircular shape, as shown in fig. 24 (fig. 2D).

In such a case, a wedge 210 composed of Polytetrafluoroethylene (PTFE) resin is used. For the wedge used, the angle θ 1 of the bonding interface with the contact surface of the wedge 210 against the first substrate 10 is 20 °, while the angle θ 2 of the bonding interface with the contact surface of the wedge 210 against the second substrate 20 is 10 °. That is, the relation 0. ltoreq. theta.2theta. ltoreq. theta.1 is satisfied (FIG. 8 or 9).

Next, the

substrate holding portions

103 and 104 in the process apparatus 600 shown in fig. 20 hold the bonded substrate stack 30. The bonded substrate stack 30 is positioned away from the initial portion opposite the nozzle 106.

By adding 400 kg-force/cm²The separation is started by the fluid (in this case water) being ejected from a nozzle having a diameter of 0.1 mm to the separation initiation portion. Thereafter, the bonded substrate stack 30 is rotated while being separated, whereby a crack grows spirally along the interface in the second porous layer near the interface between the first and second porous layers. Since the crack reaches the center of the bonded substrate stack 30, the substrate stack 30 bonded on the portion of the porous layer 12 can be completely separated into two pieces of substrates (fig. 2E). With hydrofluoric acid having 49 wt.% HF, having 30 wt.% H₂O₂The mixed solution of hydrogen peroxide and water as an etchant, selectively etching the remaining porous layer 12b remaining on the surface of the first substrate 20' after the separation (fig. 2F). With such a process, an SOI substrate as shown in fig. 2F is obtained. The outer edge of the converted single crystal Si layer 13b is present within 2 mm from the outer edge of the substrate 20 and the outer edge of the converted single crystal Si layer 13b is in close contact with the substrate 20 via the insulating layer 14b over the entire circumference.

Next, hydrofluoric acid with 49 wt% HF, with 30 wt% H₂O₂The mixed solution of hydrogen peroxide and water is used as an etchant for selective etchingThe remaining porous layer 12b remaining on the single crystal Si substrate 11 is etched (fig. 2G).

[ example 2]

In example 2, the second process apparatus 300 and the fifth process apparatus 600 are applied to a process of fabricating an SOI substrate.

The processes up to the time of forming the bonded substrate stack 30 are completed in the same procedure as in example 1 described above.

As shown in fig. 11, the outer peripheral portion of the bonded substrate stack 20 is held by the holding portion 203 in the process apparatus 300. At this time, the separation initiation portion formed in the bonded substrate stack 30 is made opposite to the wedge 220.

Motor 213 is driven tomove wedge 220 parallel to the bonding interface of bonded stack 30 and to insert wedge 220 into the beginning of the separation of bonded stack 30 by only about 1.5 mm.

In such a case, a wedge 220 composed of PTFE is used. For the wedge 220 used, the wedge 220 is withdrawn against the contact surface of the second substrate 20 only 0.5 mm (i.e., D =0.5 mm>0) from the first contact surface of the wedge 220 against the first substrate 10 (fig. 12 or 13). Thus, a crack reaching porous layer 12 as shown in fig. 6 is formed in bonded substrate stack 30. The portion at the end (outer face) of the crack is deleted to form a separation starting portion having a nearly semicircular shape, as shown in fig. 24 (fig. 2D).

Next, as in example 1, the actual separation process is completed by the process equipment 600 shown in fig. 20, and an SOI substrate is formed. Further, as in example 1, the single crystal Si substrate was reused.

[ example 3]

In example 3, the third process apparatus 400 and the fifth process apparatus 600 are applied to a process of fabricating an SOI substrate.

As shown in fig. 14, the outer peripheral portion of the bonded substrate stack 30 is held by a holding portion 203 in the process apparatus 400. At this time, the separation initiation portion formed in the bonded substrate stack 30 is made opposite to the wedge 230.

The motor 213 is driven to move the wedge 230 parallel to the bonding interface of the bonded stack 30 and to insert the wedge 230 into the bonded stack 30 only about 1.5 mm apart.

For the wedge 230 used, the first element 230a thatcollides against the first substrate 10 has a higher hardness than the second element that collides against the second substrate 20. More precisely, wedges 230 are used, consisting of PEEK for the first element 230a and rubber for the second element 230b (fig. 15 or 16).

Thus, cracks reaching the porous layer as shown in fig. 6 are formed in the bonded substrate stack 30. The portion outside the crack is deleted to form a separation starting portion having a nearly semicircular shape, as shown in fig. 20 (fig. 2D).

[ example 4]

In example 4, the fourth process apparatus 500 and the fifth process apparatus 600 are applied to a process of fabricating an SOI substrate.

As shown in fig. 17, the outer peripheral portion of the bonded substrate stack 30 is held by a holding portion 203 in the process apparatus 500. At this time, the separation initiation portion formed in the bonded substrate stack 30 is made opposite to the wedge 240.

The motor 213 is driven to move the wedge 240 parallel to the bonding interface of the bonded substrate stack 30 and to insert the wedge 240 into the bonded substrate stack 30 only about 1.5 mm apart by the initial portion when the wedge 240 is vibrated. Thus, a crack reaching porous layer 12 as shown in fig. 6 is formed in bonded substrate stack 30. The portion at the end of the crack was deleted to form a separation starting portion having a nearly semicircular shape, as shown in fig. 20 (fig. 2D).

[ example 5]

In example 5, the first process equipment 200 and the fifth process equipment 600 are used for a process of fabricating an SOI substrate.

The process until the formation of the separation start portion 60 in the bonded substrate stack 30 is completed in the same procedure as in example 1 described above.

Next, as in example 1, the process apparatus 600 shown in fig. 20 holds the bonded substrate stack 30 and rotates the bonded substrate stack 30, and ejects a fluid at a pressure of 400 kgf/cm.

After the fluid impacts toward the portion near the rotating separation initiation portion 60, the crack grows spirally to reach the center of the bonded substrate stack 30, so that the bonded substrate stack 30 is completely separated into two substrates.

The subsequent processes were the same as in example 1.

[ example 6]

In example 6, the sixth process equipment 700 is used for a process of fabricating an SOI substrate.

Next, the process apparatus 700 shown in fig. 21 holds the bonded substrate stack 30, and sprays fluid to the bonded substrate stack 30 in a stationary state at a pressure of 2000 kgf/cm to warp the end surface portion of the substrate 10 downward to form a separation start portion.

The fluid pressure is reduced to 400 kgf/cm and then the bonded substrate stack 30 is rotated to spirally grow cracks, thereby completely separating the bonded substrate stack 30 into two substrates.

The subsequent processes were the same as in example 1.

According to the present invention, for example, a combined element such as a bonded substrate stack can be appropriately separated at a separation layer such as a porous layer.

Since many apparently widely different embodiments of the invention can be made without departing from the spirit and scope thereof, it is to be understood that the invention is not limited to the specific embodiments disclosed herein, except as defined in the appended claims.

Claims

1. A composite member separating method for separating a composite member formed by bonding a first member having a separation layer and a conversion layer on the separation layer to a second member at a position different from a bonding interface between the first member and the second member, comprising:

a pre-separation step comprising the step of applying an asymmetric force to the end portion of the composite member relative to the bonding interface to form a crack in the composite member extending from the first member surface through the conversion layer to the separation layer; and

an actual separation step, which comprises a step of growing a crack along the separation layer.

2. The method of claim 1, wherein the step of actually separating comprises growing a crack substantially within the separation layer or within an interface of the separation layer.

3. The method of claim 1, wherein the conversion layer comprises an insulating layer.

4. The method of claim 1, wherein the conversion layer comprises an insulating layer and a semiconductor layer.

5. The method according to claim 1, wherein the separation layer comprises a porous layer formed by anodizing.

6. The method according to claim 1, wherein the separation layer comprises an implanted layer formed by ion implantation.

7. The method of claim 1, wherein the pre-separating step comprises applying an asymmetric force to the end portion of the composite member relative to the bonding interface using at least one of a solid wedge and an asymmetric substrate clamping mechanism.

8. The method of claim 7, wherein the wedge has a pair of asymmetrically sloped contact surfaces.

9. The method of claim 7, wherein the wedge has a first contact surface against the first element and a second contact surface against the second element, and the first contact surface makes a larger angle with the bonding interface than the second contact surface makes with the bonding interface.

10. The method of claim 7, wherein the pre-separating step comprises moving the wedge parallel to the bonding interface to insert the wedge into the end face portion of the composite member.

11. The method according to claim 7, wherein the wedge has a first contact surface against the first element and a second contact surface against the second element, and the wedge has a structure in which the first contact surface is in contact with the combination element before the second contact surface is in contact with the combination element in a condition in which the wedge is inserted into the combination element.

12. The method according to claim 7, wherein the wedge has a first contact part abutting the first element and a second contact part abutting the second element, the first contact part having a higher hardness than the hardness of the second contact part.

13. The method of claim 7, wherein the pre-separating step comprises applying an asymmetric vibration to the wedge relative to the bonding interface.

14. The method of claim 7, wherein the wedges are comprised of resin.

15. The method of claim 1, wherein the step of physically separating comprises injecting a fluid into the crack to propagate the crack.

16. The method of claim 15, wherein the fluid is a hydrostatic fluid or an ejection fluid.

17. The method of claim 1, wherein the pre-separating step includes forming the crack without rotating the composite member and the actual separating step includes growing the crack while rotating the composite member.

18. The method of claim 1, wherein the pre-separating step includes inhibiting warpage of the second component and cracking the second component such that the second component has less warpage than the first component.

19. The method of claim 1, wherein the pre-separating step includes clamping the composite member using a substrate clamping portion having features that inhibit warping of the second member such that the warping of the second member is less than the warping of the first member.

20. A method of manufacturing a thin film, comprising the step of converting a conversion layer in a first element on a surface side of a separation layer to a second element using the separation method of claim 1.

21. A method of making a film comprising:

a step of bonding a first element having a separation layer and a conversion layer on the separation layer to a second element to form a combined element; and

a separation step of separating the composite member at a position different from the bonding interface between the first member and the second member,

wherein the separation step comprises

A first step of applying a force to the end face portion of the composite element which is asymmetric with respect to the bonding interface to form a crack in the composite element extending from the first element surface through the conversion layer to the separation layer, and

a second step of growing a crack along the separation layer.

22. A method according to claim 21, wherein the second step comprises growing the crack substantially within the separation layer or within an interface of the separation layer.

23. The method of claim 21, wherein theconversion layer comprises an insulating layer.

24. The method of claim 21, wherein the conversion layer comprises an insulating layer and a semiconductor layer.

25. A method according to claim 21 wherein the separation layer comprises a porous layer formed by anodisation.

26. The method of claim 21, wherein the separation layer comprises an implanted layer formed by ion implantation.

27. The method of claim 21, wherein the first step comprises applying an asymmetric force to the face portion of the composite member relative to the bonding interface using at least one of a solid wedge and an asymmetric substrate clamping mechanism.

28. The method of claim 27, wherein the wedge has a pair of asymmetrically sloped contact surfaces.

29. The method of claim 27, wherein the wedge has a first contact surface for contacting the first component and a second contact surface for contacting the second component, and the first contact surface makes a larger angle with the bonding interface than the second contact surface makes with the bonding interface.

30. A method according to claim 27, wherein the first step comprises moving the wedge parallel to the bonding interface to insert the wedge into the end face portion of the composite member.

31. The method according to claim 27, wherein the wedge has a first contact part against the first element and a second contact part against the second element, and the wedge has a configuration in which the first contact part combines the elements before the second contact part combines the elements in a condition in which the wedge is inserted into the combined elements.

32. The method according to claim 27, wherein the wedge has a first contact part abutting the first element and a second contact part abutting the second element, the first contact part having a higher hardness than the hardness of the second contact part.

33. The method of claim 27, wherein the first step comprises applying an asymmetric vibration to the wedge relative to the bonding interface.

34. The method of claim 27, wherein the wedges are comprised of resin.

35. A method according to claim 2, wherein the second step comprises injecting a fluid into the crack to grow the crack.

36. The method of claim 35, wherein the fluid is a hydrostatic fluid or an ejection fluid.

37. The method according to claim 21, wherein the first step comprises forming the crack without rotating the combination element, and the second step comprises growing the crack while rotating the combination element.

38. The method according to claim 21, wherein the first step includes forming cracks while suppressing warpage of the second member so that the warpage of the second member is smaller than that of the first member.

39. The method according to claim 38, wherein the first step comprises holding the combined component with a substrate holding portion having a member for suppressing warping of the second component so that the warping of the second component is smaller than the warping of the first component.

40. The method according to claim 21, further comprising a step of forming a non-porous semiconductor layer and an insulating layer covering the porous layer after forming the porous layer as a separation layer on the surface of the substrate, thereby forming the first element.

41. The method according to claim 21, further comprising a step of forming an ion-implanted layer serving as a separation layer in the substrate by implanting ions from a surface side of the substrate, thereby forming the first element.

42. The method of claim 21, further comprising the step of heat treating the composite member in an oxidizing atmosphere.

43. The method of claim 21, wherein

The above method further comprises the step of heat-treating the composite member in an oxidizing atmosphere to form an oxide film on the surface of the composite member, and

a crack extending from the first element to the separation layer traverses the peroxide membrane.

44. The method of claim 21, wherein the thin film is an SOI layer formed on an insulating surface.

45. A composite member separating apparatus for separating a composite member formed by bonding a first member having a separation layer and a conversion layer on the separation layer to a second member at a position different from a bonding interface between the first member and the second member, comprising:

a pre-separation mechanism for applying an asymmetric force to the end face portion of the composite element relative to the bonding interface to form a crack in the composite element extending from the first element surface through the conversion layer to the separation layer; and

actual separation mechanism for growing cracks along the separation layer.